© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 18-GEMMASCOPE-RES-16K
    6. 9672
    Prev
    Next
    INDEX
    Explanations

    assertions about required conditions and adjustments

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    followed by "to" indicating a requirement/need

    np_acts-logits-general · gemini-2.0-flash

    must be fulfilled

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 18-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_18/width_16k/average_l0_74
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.18.hook_resid_post
    Hook Layer
    18
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     dandy
    -0.50
    UnusedPrivate
    -0.50
    isling
    -0.50
    brainly
    -0.47
    getDefault
    -0.47
    oczes
    -0.47
    UNUSED
    -0.46
    fxml
    -0.46
    __":
    
    -0.46
    
    -0.46
    POSITIVE LOGITS
    才行
    1.80
    才能
    1.65
    才可以
    1.43
    才有
    1.20
    才會
    1.15
    才会
    1.11
    しないと
    1.07
     must
    0.96
     才
    0.95
     inorder
    0.92
    Activations Density 0.639%

    No Known Activations